Red Wine Quality Analysis
Li Cai
March 13, 2017
In this project, I will explore the dataset about red wind quality to find which chemical properties influence it.
Summary Statistics
## 'data.frame': 1599 obs. of 13 variables:
## $ X : int 1 2 3 4 5 6 7 8 9 10 ...
## $ fixed.acidity : num 7.4 7.8 7.8 11.2 7.4 7.4 7.9 7.3 7.8 7.5 ...
## $ volatile.acidity : num 0.7 0.88 0.76 0.28 0.7 0.66 0.6 0.65 0.58 0.5 ...
## $ citric.acid : num 0 0 0.04 0.56 0 0 0.06 0 0.02 0.36 ...
## $ residual.sugar : num 1.9 2.6 2.3 1.9 1.9 1.8 1.6 1.2 2 6.1 ...
## $ chlorides : num 0.076 0.098 0.092 0.075 0.076 0.075 0.069 0.065 0.073 0.071 ...
## $ free.sulfur.dioxide : num 11 25 15 17 11 13 15 15 9 17 ...
## $ total.sulfur.dioxide: num 34 67 54 60 34 40 59 21 18 102 ...
## $ density : num 0.998 0.997 0.997 0.998 0.998 ...
## $ pH : num 3.51 3.2 3.26 3.16 3.51 3.51 3.3 3.39 3.36 3.35 ...
## $ sulphates : num 0.56 0.68 0.65 0.58 0.56 0.56 0.46 0.47 0.57 0.8 ...
## $ alcohol : num 9.4 9.8 9.8 9.8 9.4 9.4 9.4 10 9.5 10.5 ...
## $ quality : int 5 5 5 6 5 5 5 7 7 5 ...The dataset has 1599 ovservations of 13 variables, and varicable X can be simply seen as index, so I’d like to remove this column.
Now, I have 12 variables. Since the most important part is quality, it would be necessary to see the basic statistics on it.
## fixed.acidity volatile.acidity citric.acid residual.sugar
## Min. : 4.60 Min. :0.1200 Min. :0.000 Min. : 0.900
## 1st Qu.: 7.10 1st Qu.:0.3900 1st Qu.:0.090 1st Qu.: 1.900
## Median : 7.90 Median :0.5200 Median :0.260 Median : 2.200
## Mean : 8.32 Mean :0.5278 Mean :0.271 Mean : 2.539
## 3rd Qu.: 9.20 3rd Qu.:0.6400 3rd Qu.:0.420 3rd Qu.: 2.600
## Max. :15.90 Max. :1.5800 Max. :1.000 Max. :15.500
## chlorides free.sulfur.dioxide total.sulfur.dioxide
## Min. :0.01200 Min. : 1.00 Min. : 6.00
## 1st Qu.:0.07000 1st Qu.: 7.00 1st Qu.: 22.00
## Median :0.07900 Median :14.00 Median : 38.00
## Mean :0.08747 Mean :15.87 Mean : 46.47
## 3rd Qu.:0.09000 3rd Qu.:21.00 3rd Qu.: 62.00
## Max. :0.61100 Max. :72.00 Max. :289.00
## density pH sulphates alcohol
## Min. :0.9901 Min. :2.740 Min. :0.3300 Min. : 8.40
## 1st Qu.:0.9956 1st Qu.:3.210 1st Qu.:0.5500 1st Qu.: 9.50
## Median :0.9968 Median :3.310 Median :0.6200 Median :10.20
## Mean :0.9967 Mean :3.311 Mean :0.6581 Mean :10.42
## 3rd Qu.:0.9978 3rd Qu.:3.400 3rd Qu.:0.7300 3rd Qu.:11.10
## Max. :1.0037 Max. :4.010 Max. :2.0000 Max. :14.90
## quality
## Min. :3.000
## 1st Qu.:5.000
## Median :6.000
## Mean :5.636
## 3rd Qu.:6.000
## Max. :8.000Since the most important part is quality, it would be necessary to see the basic statistics on it. From the dataset documentation, the wine quality is rated on 0-10, actually, the quality values on this data set has the lowest rate 3 and highest rate 8, with a median of 6 and a mean of 5.6.
Univariate Plots Section
At first, I will plot 12 simple historams of all 12 variables to see how they distribute.
These 12 histograms reveal that density and PH are normal disributed but fixed.acidity, volatile.acidity, total.sulfur.dioxide, and sulphates seem to be long-tailed.
On log10 scale, fixed.acidity, volatile.acidity, total.sulfur.dioxide, and sulphates appear to be normal distributed, although they are still with some outliers.
In this dataset, quality rate is from 3-8, and the large majority of the red wines are rated from 5-7. It seems to make sense that create a new rateing including bad(0-4), avreage(5-7) and good(8-10).
## bad average good
## 63 1518 18The rating result is that more than 90% of red wine are ‘average’.
Plot it.
It’s more straight to see taht most wine are in average rating.
Univariate Analysis
What is the structure of your dataset?
Now, there are 1599 obsevations of 13 variables in the data set, short discriptions of 13 variables are as follows:
- fixed acidity: most acids involved with wine or fixed or nonvolatile (do not evaporate readily)
- volatile acidity: the amount of acetic acid in wine
- citric acid: add ‘freshness’ and flavor to wines
- residual sugar: the amount of sugar(gram/liter) remaining after fermentation stops
- chlorides: the amount of salt in the wine
- free sulfur dioxide: the free form of SO2 exists in equilibrium between molecular SO2 (as a dissolved gas) and bisulfite ion
- total sulfur dioxide: amount of free and bound forms of S02
- density: the density of water is close to that of water depending on the percent alcohol and sugar content
- pH: describes how acidic or basic a wine is on a scale from 0 (very acidic) to 14 (very basic); most wines are between 3-4 on the pH scale
- sulphates: a wine additive which can contribute to sulfur dioxide gas (S02) levels, wich acts as an antimicrobial and antioxidant
- alcohol: the percent alcohol content of the wine
- quality: score between 0 and 10
- rating: rated on bad(0-4), average(5-7), good(8-10)
What is/are the main feature(s) of interest in your dataset?
Quality of the red wine is the main featur in the dataset and I will also take a look at how other variables would influence the quality of the wine.
What other features in the dataset do you think will help support your
investigation into your feature(s) of interest?
The Variable residual.sugar seems to be an intersting one, and I’ll explore other variables like alcohol and PH.
Did you create any new variables from existing variables in the dataset?
Yes, rating is the variable I created.
Of the features you investigated, were there any unusual distributions?
Did you perform any operations on the data to tidy, adjust, or change the form
of the data? If so, why did you do this?
As I meantioned above, density and PH are normal disributed but fixed.acidity, volatile.acidity, total.sulfur.dioxide, and sulphates seem to be long-tailed. Also, I remove X from orininal dataset, because it just the index.
Bivariate Plots Section
First, I’d like to plot the correlation of all varianles against each other.
Clearly, there are some strong correlations between some variables such as total.sulfur.dioxide and free.sulfur.dioxide, volatile.acidity and fixed.acidity, total.sulfur.dioxide and fixed.acidity, fixed.acidity and PH.
Also, this plots tells that quality has higher corelations with alcohol, sulphates, density, total.sulfur.dioxide, chlorides, citric.acid, volatile.acidity and fixed.acidity than other variables.
Create boxplots of these variables vs. quality can see how they affect the quality of the red wine.
It seems like that higher quality red wine has higher fixed.acidity, but it’s not obvious.
It’s clear that higher quality red wine has lower volatile.acidity.
It’s clear that higher quality red wine has higher citric.acid.
It seems like that higher quality red wine has lower chlorides.
Cannot see specific relationship between total.sulfur.dioxide and quality.
It shows that higher quality red wine has lower density.
It shows that higher quality red wine has higher sulphates.
It shows that higher quality red wine has higher alcohols.
After explored all boxplots, there are no obvious relayionship between total.sulfur.dioxide and quality and I found that a good wine seems has the following characteristic:
- higer fixed.acidity, citric.acid, sulphates, and alcohol
- lower volatile.acidity, chlorides and density
And then I will calculate the correlation for each of these seven variable against quality.
## fixed.acidity citric.acid sulphates alcohol
## 0.1240516 0.2263725 0.2513971 0.4761663
## volatile.acidity chlorides density
## -0.3905578 -0.1289066 -0.1749192The result shows that the correlations between quality and fixed.acidity, chlorides, density are lower than 0.2, it’s pretty samll. And then I’ll plot the rest four variables vs. quality with removing some outliers.
It also will be interesting to see the correlations between variables about acid, and acidity against PH. First, calculate correlations and then create the scattor plots.
## fixed.acidity_citric.acid citric.acid_volatile.acidity
## 0.6717034 -0.5524957
## volatile.acidity_fixed.acidity PH_total.acidity
## -0.2561309 -0.6834838
Obviously, these variables has high correaltions. citric.acid and fixed.acidity, fixed.acidity and volatile.acidity have highly positive correlations, citric.acid and volatile.acidity, total.acidity and PH have highly negative correlations.
Bivariate Analysis
Talk about some of the relationships you observed in this part of the
investigation. How did the feature(s) of interest vary with other features in
the dataset? From the boxplot, I found some trends that a good red wine has. After the caluculation of eight variables against the quality of red wine and create scatterplot, four rariables were removed, and the result is a good wine seems has the following characteristic:
- higer fixed.acidity, citric.acid, sulphates, and alcohol
- lower volatile.acidity, chlorides and density
Did you observe any interesting relationships between the other features
(not the main feature(s) of interest)?
I observed the correlations between acid variables, and ph vs. total.acidity, results are as follows:
- fixed.acidity & citric.acid : 0.6717034
- citric.acid & volatile.acidity: -0.5524957
- volatile.acidity & fixed.acidity: -0.2561309
- PH & total.acidity: -0.6834838
The most important relationship is between quality and alcohol, alcohol have the highest correlation(0.4761663) with quality of red wine.
What was the strongest relationship you found?
The strongest relationship I found is PH vs.total.acidity, which is more than -0.68, and the second is correlation between fixed.acidity and citric.acid(0.6717034).
Multivariate Plots Section
As mentioned above, alcohol has the highest correlation with the quality of re wines, so I will create multivariate plots that are alohol vs. other three variables and rating. For the rating, I will remove average rating before plot because average has the most quantities, remove it can make the plot more obvious to find the differences between ‘good’ and ‘bad’.
For a ‘good’ red wine, it has higher alcohol and lower volatile.acidity, and it has more positive correlation between alcohol and volatile.acidity.
For a ‘good’ red wine, it has higher alcohol but is not clear for volatile.acidity, and it has more negative correlation between alcohol and citric.acid.
For a ‘good’ red wine, it has higher alcohol and higer sulphates, and it has more negative correlation between alcohol and sulphates.
Multivariate Analysis
Talk about some of the relationships you observed in this part of the
investigation. Were there features that strengthened each other in terms of
looking at your feature(s) of interest?
Only one thing can be comfirmed, that is good red wine have higer alcohol. Other three factors have limit influence on quality of red wine.
Were there any interesting or surprising interactions between features?
The cititric.acid for both good or bad red wind doesn’t have clear differences.
Final Plots and Summary
Plot One
Description One
This histogram clearly reveals that most red wind is on average rating, and most of these ‘avearge’ red wines have quality 5 or 6.
Plot Two
Description Two
These boxplots demonstrate the effect of alcohol on red wine quality, that is, higer quality red wine has higer alcohol, some outliers doesn’t show this relationship.
Plot Three
Description Three
After remove the avearge rating wind data, this scatterplot shows the realationship between alcohol and the quality of red wine again. Additionaly, the realationship of vlotile.acidity and quality can be seen from this plot, higer quality red wine has lower volatile.acidity. And the good wine has more positive correlation between alcohol and volatile.acidity.
Reflection
The object of this exploratory data analysis is to find out which chemical properties would afftect the quality of red wines.In order to obverse the quality more directly, I divided the quality into new three rating: bad, average and good. I plotted and calculated the correlations between quality and the variables. However, after all these anaysis, I found none of the correlations were above 0.7. Aalcohol is the most important facor that influence the quality of red wines, the acidity also affect the quality to some extent. I hvae to say the measure of red wine quality is subjective, which means the data analysis is not enough to reveal all factors and to rate a res wine.
In this dataset, more than 80% red wines are ratedas 5 or 6, this limitation makes that there is not enough data to analyze good wine’s factors. In furter analysis, a dataset has more obeervations woll be preferred. However,